Files
EcgSynKit/ecgsyn-1.0.0/paper/node5.html
2024-11-04 10:48:09 -06:00

249 lines
9.4 KiB
HTML

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
<!--Converted with jLaTeX2HTML 2002 (1.62) JA patch-1.4
patched version by: Kenshi Muto, Debian Project.
LaTeX2HTML 2002 (1.62),
original version by: Nikos Drakos, CBLU, University of Leeds
* revised and updated by: Marcus Hennecke, Ross Moore, Herb Swan
* with significant contributions from:
Jens Lippmann, Marek Rouchal, Martin Wilck and others -->
<HTML>
<HEAD>
<TITLE>Results</TITLE>
<META NAME="description" CONTENT="Results">
<META NAME="keywords" CONTENT="ecgpqrst">
<META NAME="resource-type" CONTENT="document">
<META NAME="distribution" CONTENT="global">
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="jLaTeX2HTML v2002 JA patch-1.4">
<META HTTP-EQUIV="Content-Style-Type" CONTENT="text/css">
<LINK REL="STYLESHEET" HREF="ecgpqrst.css">
<LINK REL="next" HREF="node6.html">
<LINK REL="previous" HREF="node4.html">
<LINK REL="up" HREF="index.shtml">
<LINK REL="next" HREF="node6.html">
</HEAD>
<BODY >
<!--Navigation Panel-->
<A NAME="tex2html75"
HREF="node6.html">
<IMG WIDTH="37" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="next"
SRC="file:/usr/share/latex2html/icons/next.png"></A>
<A NAME="tex2html73"
HREF="index.shtml">
<IMG WIDTH="26" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="up"
SRC="file:/usr/share/latex2html/icons/up.png"></A>
<A NAME="tex2html67"
HREF="node4.html">
<IMG WIDTH="63" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="previous"
SRC="file:/usr/share/latex2html/icons/prev.png"></A>
<BR>
<B> Next:</B> <A NAME="tex2html76"
HREF="node6.html">Conclusions</A>
<B> Up:</B> <A NAME="tex2html74"
HREF="index.shtml">A dynamical model for</A>
<B> Previous:</B> <A NAME="tex2html68"
HREF="node4.html">The dynamical model</A>
<BR>
<BR>
<!--End of Navigation Panel-->
<H1><A NAME="SECTION00050000000000000000"></A>
<A NAME="s:results"></A>
<BR>
Results
</H1>
The synthetic ECG (Fig. <A HREF="node5.html#f:ecgsynthetic">5</A>) illustrates
the modulation of the QRS-complex due to RSA.
Observational uncertainty is incorporated by adding normally
distributed measurement errors with mean zero
and standard deviation 0.025 mV (Fig. <A HREF="node5.html#f:ecgcomparison">6</A>a),
yielding a similar signal to a segment of real ECG from a normal human
(Fig. <A HREF="node5.html#f:ecgcomparison">6</A>b).
In order to illustrate the
dynamics of the RR-intervals obtained from this synthetic ECG, peak detection
was used to identify the times of the R-peaks.
In the noise-free case, a simple algorithm which looks for local maxima
within a small window is sufficient. For ECGs with noise and artefacts it
may be necessary to use more complicated methods [<A
HREF="node8.html#pan85">2</A>,<A
HREF="node8.html#kaplan91">3</A>].
A comparison between the continuous process with power spectrum <IMG
WIDTH="37" HEIGHT="31" ALIGN="MIDDLE" BORDER="0"
SRC="img2.png"
ALT="$S(f)$">
given by (<A HREF="node4.html#e:Sf">3</A>) and the piecewise constant
reconstruction of the RR-process obtained from the R-peak detection
(Fig. <A HREF="node5.html#f:rrinout">7</A>) illustrates the measurement errors that
arise when computing heart rate variability statistics from
RR-intervals.
The RR-intervals (Fig. <A HREF="node5.html#f:rrsynthetic">8</A>a) and corresponding
instantaneous heart rate (Fig. <A HREF="node5.html#f:rrsynthetic">8</A>b)
in units of beats per minute (bpm)
for a mean of 60 bpm and standard deviation of 5 bpm
display variability due to both RSA and Mayer waves.
A spectral estimation technique
for unevenly sampled time series, the Lomb periodogram
[<A
HREF="node8.html#press92">15</A>,<A
HREF="node8.html#laguna98">16</A>], was used to calculate the power
spectrum (Fig. <A HREF="node5.html#f:rrsynthetic">8</A>c) from the RR tachogram,
derived from 5 minutes of data as recommended
by [<A
HREF="node8.html#malik95">7</A>,<A
HREF="node8.html#eursoccard96">10</A>].
Despite the loss of information in going from the continuous process to the
piecewise constant reconstruction, a comparison between Fig. <A HREF="node4.html#f:Sf">4</A> and
Fig. <A HREF="node5.html#f:rrsynthetic">8</A>c illustrates that it is still
possible to obtain a reasonable estimate of the power spectrum.
<DIV ALIGN="CENTER"><A NAME="f:ecgsynthetic"></A><A NAME="168"></A>
<TABLE>
<CAPTION ALIGN="BOTTOM"><STRONG>Figure 5:</STRONG>
ECG generated by dynamical model: (a) 10 seconds and (b) 50 seconds.</CAPTION>
<TR><TD><IMG
WIDTH="352" HEIGHT="275" BORDER="0"
SRC="img63.png"
ALT="\begin{figure}
\centerline{\psfig{file=ecgsynthetic.eps,width=7.75cm}}
\end{figure}"></TD></TR>
</TABLE>
</DIV>
<DIV ALIGN="CENTER"><A NAME="f:ecgcomparison"></A><A NAME="173"></A>
<TABLE>
<CAPTION ALIGN="BOTTOM"><STRONG>Figure 6:</STRONG>
Comparison between (a) synthetic ECG with additive normally
distributed measurement errors and (b) real ECG signal from a normal human.</CAPTION>
<TR><TD><IMG
WIDTH="352" HEIGHT="275" BORDER="0"
SRC="img64.png"
ALT="\begin{figure}
\centerline{\psfig{file=ecgcomparison.eps,width=7.75cm}}
\end{figure}"></TD></TR>
</TABLE>
</DIV>
<DIV ALIGN="CENTER"><A NAME="f:rrinout"></A><A NAME="296"></A>
<TABLE>
<CAPTION ALIGN="BOTTOM"><STRONG>Figure 7:</STRONG>
Reconstruction of RR-process from R-peak detection:
the underlying RR-process generated using (<A HREF="node4.html#e:Sf">3</A>) (black line)
and the RR-interval time series obtained using R-peak
detection of the synthetic ECG (grey line).</CAPTION>
<TR><TD><IMG
WIDTH="352" HEIGHT="275" BORDER="0"
SRC="img65.png"
ALT="\begin{figure}
\centerline{\psfig{file=rrinout.eps,width=7.75cm}}
\end{figure}"></TD></TR>
</TABLE>
</DIV>
<DIV ALIGN="CENTER"><A NAME="f:rrsynthetic"></A><A NAME="298"></A>
<TABLE>
<CAPTION ALIGN="BOTTOM"><STRONG>Figure 8:</STRONG>
Analysis of RR-intervals from R-peak detection of the ECG signal
generated by the dynamical model (<A HREF="node4.html#e:pqrst">1</A>) with mean heart rate 60 bpm
and standard deviation 5 bpm: (a) RR-intervals,
(b) instantaneous heart rate and (c) power spectrum of the RR-intervals.
Note the two active frequencies belonging to RSA (0.25 Hz) and Mayer
waves (0.1 Hz).</CAPTION>
<TR><TD><IMG
WIDTH="352" HEIGHT="272" BORDER="0"
SRC="img66.png"
ALT="\begin{figure}
\centerline{\psfig{file=rrsynthetic.eps,width=7.75cm}}
\end{figure}"></TD></TR>
</TABLE>
</DIV>
An increase in the RR-interval implies that the trajectory has more time to
get pushed into the peak and trough given by the R and S events.
This is reflected by the strong correlation between the RR-intervals and the
RS-amplitude as shown in Fig. <A HREF="node5.html#f:rsrr">9</A>. A technique for deriving a measure
of the rate of respiration
from the ECG has been proposed [<A
HREF="node8.html#moody85">5</A>,<A
HREF="node8.html#moody86">6</A>].
This ECG-derived respiratory signal (EDR) is of clinical use in
situations where the ECG, but not respiration, is recorded.
The synthetic ECG provides a means of testing the robustness of such
techniques against noise and the effects of different sampling frequencies.
<DIV ALIGN="CENTER"><A NAME="f:rsrr"></A><A NAME="190"></A>
<TABLE>
<CAPTION ALIGN="BOTTOM"><STRONG>Figure 9:</STRONG>
RS-amplitudes versus RR-intervals for the synthetic ECG.</CAPTION>
<TR><TD><IMG
WIDTH="351" HEIGHT="275" BORDER="0"
SRC="img67.png"
ALT="\begin{figure}
\centerline{\psfig{file=rsrr.eps,width=7.75cm}}
\end{figure}"></TD></TR>
</TABLE>
</DIV>
As a consequence of constructing the model with a variable angular frequency
<IMG
WIDTH="33" HEIGHT="31" ALIGN="MIDDLE" BORDER="0"
SRC="img61.png"
ALT="$\omega(t)$">, the time taken to
move from the Q event to the T event, known as the QT-interval, varies with
the RR-interval on a beat-to-beat basis.
The relationship between the QT-interval and the
RR-interval is linear as shown in Fig. <A HREF="node5.html#f:qtrr">10</A>.
Such a linear relationship has been reported for real ECGs and
has been used to calculate a corrected QT-interval [<A
HREF="node8.html#davey99">4</A>].
It is interesting that this relationship is a direct consequence of the
model. Furthermore it may be possible to use the model to assess how much of
the variation in the QT-interval is due to RR-interval variability so that
this effect can be factored out.
<DIV ALIGN="CENTER"><A NAME="f:qtrr"></A><A NAME="197"></A>
<TABLE>
<CAPTION ALIGN="BOTTOM"><STRONG>Figure 10:</STRONG>
QT-intervals versus RR-intervals for the synthetic ECG.</CAPTION>
<TR><TD><IMG
WIDTH="352" HEIGHT="275" BORDER="0"
SRC="img68.png"
ALT="\begin{figure}
\centerline{\psfig{file=qtrr.eps,width=7.75cm}}
\end{figure}"></TD></TR>
</TABLE>
</DIV>
<HR>
<!--Navigation Panel-->
<A NAME="tex2html75"
HREF="node6.html">
<IMG WIDTH="37" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="next"
SRC="file:/usr/share/latex2html/icons/next.png"></A>
<A NAME="tex2html73"
HREF="index.shtml">
<IMG WIDTH="26" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="up"
SRC="file:/usr/share/latex2html/icons/up.png"></A>
<A NAME="tex2html67"
HREF="node4.html">
<IMG WIDTH="63" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="previous"
SRC="file:/usr/share/latex2html/icons/prev.png"></A>
<BR>
<B> Next:</B> <A NAME="tex2html76"
HREF="node6.html">Conclusions</A>
<B> Up:</B> <A NAME="tex2html74"
HREF="index.shtml">A dynamical model for</A>
<B> Previous:</B> <A NAME="tex2html68"
HREF="node4.html">The dynamical model</A>
<!--End of Navigation Panel-->
<ADDRESS>
2003-10-08
</ADDRESS>
</BODY>
</HTML>